Mitigating the forward guidance puzzle: Inattention, credibility, finite planning horizons and learning

This paper develops a simple, consistent methodology for generating empirically realistic forward guidance simulations using existing macroeconomic models by modifying expectations about policy announcements. The main advantage of our method lies in the exact preservation of all other shock transmissions. We describe four scenarios regarding how agents incorporate information about future interest rate announcements: "inattention", "credibility", "finite planning horizon", and "learning". The methodology consists of describing a single loading matrix that augments the equilibrium decision rules and can be applied to any standard DSGE, including large-scale policy-institution models. Finally, we provide conditions under which the forward guidance puzzle is resolved.